The massive problems of the patent system really started getting renewed attention between 2002 and 2004 or so, highlighted by the publication of the book Innovation and Its Discontents: How Our Broken Patent System Is Endangering Innovation and Progress, and What to Do About It
by Adam Jaffe and Josh Lerner. By that point, the combination of two key events in the late 90s was clearly being felt on the patents system. First, and most importantly, was the impact of the State Street decision
that announced to the world that the courts considered software and business method patents legal. Also important was the 1999 publication of Rembrandts in the Attic: Unlocking the Hidden Value of Patents
by Kevin Rivette and David Kline, which led patent lawyers and tech companies alike to suddenly both ramp up their patenting, but also to look to sell off "unused" patents to companies (lawyers) who did nothing but threaten and sue over them. Suddenly, patent trolls became a big, big issue.
Around the time of the Jaffe and Lerner book, the USPTO seemed to actually take much of the criticism to heart. One big part of Jaffe and Lerner's criticism was the simple fact that patent examiners had significant incentives to approve patents, and almost none to reject patents. That is, the metrics by which they were measured included the rate of how many patent applications they processed. But, since there is no such thing as a truly final rejection of a patent, people would just keep asking the USPTO to look at their application again. Each time an examiner had to do this, their "rate" would decline, since they'd be spending even more time on the same old patent application. But approving
a patent got it off your plate and let the court system sort out any mess. However, after the book was published, the USPTO actually seemed to pay attention and changed its internal incentives a bit to push for high quality approvals. Not surprisingly, this meant that the approval rate dropped. But, since there was more demand for bogus patents to sue over, more people appealed the rejections and the backlog grew.
Patent system lovers started whining about the "backlog," but what they were really pissed off about was the fact that their bogus patents weren't getting approved. Unfortunately, their message resonated with the new regime of the Obama administration, mainly Commerce Dept. boss, Gary Locke, and head of the USPTO, David Kappos. Back in 2010, we noted that the USPTO had shifted back to approving "pretty much anything"
and had clearly decreased their quality standards in an effort to rush through the backlog. Not surprisingly, in stating this, we were attacked mercilessly by patent system supporters, who insisted that we were crazy, and the truth was that David Kappos had found some magic elixir that made all USPTO agents super efficient (or something like that -- their actual explanations were not much more coherent). No matter what, they insisted that it was entirely possible to massively ramp up the number of approvals, decrease the backlog and not decrease patent quality.
Needless to say, we've been skeptical that this was possible.
And now the data is in, suggesting we were absolutely right all along. A new study done by Chris Cotropia and Cecil Quillen of the University of Richmond and independent researcher Ogden Webster used information obtained via FOIA requests to delve into what was really going on in the patent office
(link to a great summary of the research by Tim Lee). The key issue, is (once again) the fact that patents are never truly rejected in full, and the people applying for patents just keep on trying again and again until someone in the USPTO approves it. However, the USPTO, to hide some of this, counts some of those "rejections" that eventually get approved as "rejections" to artificially deflate the actual "approval rate" of patent applications.
When the researchers corrected for all of this, they found that the actual patent approval rate in 2012 was almost 90% of all patents eventually get approved. 90%
! That's about where it was in 2004 and 2005 (as discussed above), though in 2001 it actually came close to 100%
! However, as noted above, by the second half of 00's corrections had been put in place and the approval rate had declined to under 70% in 2009 -- meaning that the USPTO was actually rejecting bad patents. But over the past three years, we've shot right back up. And it's clear that if the approval rate is much higher, the USPTO is approving many, many more bad patents.
In fact, it's likely that the story is even worse than before. Back in 2004 and 2005 when the approval rates were similar, it was really before the public was aware of just how bad the patent troll problem was, so you had many fewer people trying to get their own bad patents to troll over. In the past five years or so that has changed quite a bit. So the number of applications
has shot up
massively as well. In 2004 there were 382,139 applications. By 2011 that had shot up by 50% to 576,763.
I don't think anyone thinks that we suddenly became 50% more inventive between 2004 and 2011. No, the truth is that people were suddenly flooding the USPTO with highly questionable patent applications on broad and vague concepts, hoping to get a lottery ticket to shake down actual innovators. And, the USPTO under David Kappos complied, granting nearly all of them. Incredible.
When Thomas Jefferson put together the first patent system -- after being quite skeptical that patents could actually be a good thing -- he was quite careful to note that patents should only be granted in the rarest of circumstances, since such a monopoly could do a lot more harm than good. And yet, today, we encourage tons of people to send in any old bogus idea, and the USPTO has turned into little more than a rubber stamp of approval, allowing patent holders to shake down tons of people and companies, knowing that many will pay up rather than fight, and then leaving the few cases where someone fights back to be handled by the courts (who seem ignorant of the game being played).
The end result is a true disaster for actual innovation and the economy. We should all be able to agree that bad patents are not a good thing. And the USPTO is, undoubtedly, approving tons of awful patents when its true approval rate is hovering around 90%.